Mastering Generative AI: A Student's Guide to Tools, Ethics, and Evaluation

Generative AI represents a sophisticated category of artificial intelligence designed to create novel content. This can encompass a wide range of outputs, including text, images, and even video. When users interact with Generative AI tools like ChatGPT or Copilot by inputting questions or commands (known as prompts), these tools respond by generating text that mimics human writing. The underlying mechanism involves training these AI models on extensive datasets, enabling them to identify patterns in language and predict the most probable sequences of words that should logically follow one another.

It is vital to understand that most Generative AI tools are not search engines. Search engines like Google or library databases are designed to retrieve existing information and content, such as web pages, articles, and videos. In contrast, Generative AI tools create original content (text, images) by employing algorithms to predict the most probable word sequences following a user's input. Generative AI does not produce a single, fixed piece of authored work; its outputs can vary even with identical prompts, meaning the information generated may differ each time. Furthermore, Generative AI tools do not always possess up-to-date information, as they may be trained on static datasets that quickly become outdated. A significant limitation is that many Generative AI tools do not consistently cite their sources. Crucially, most Generative AI tools can generate inaccurate information, including fabricated references or citations. OpenAI, the developer of ChatGPT, explicitly states in its Terms of Use: 'Output may not always be accurate. You should not rely on Output from our Services as a sole source of truth or factual information, or as a substitute for professional advice.' Therefore, it is imperative to cross-reference AI-generated results with other reliable sources to avoid incorporating 'AI hallucinations' into your work.

Several powerful Generative AI tools are available that can handle a variety of tasks, including text generation, summarization, internet searches, and image creation. These versatile chatbot tools include Anthropic's Claude, Google's Gemini, Microsoft's Copilot, and OpenAI's ChatGPT. It is essential to reiterate the importance of checking with your instructors for their specific rules and guidelines regarding the use of these Generative AI tools for homework or papers. Adhering to these guidelines is paramount for academic integrity.

Given the nature of Generative AI, a critical approach to evaluating and verifying its output is essential. As highlighted, these tools can produce inaccurate information, including fabricated facts, references, or citations. The Terms of Use for services like OpenAI's ChatGPT explicitly caution users against relying on AI output as a sole source of truth or factual information. Therefore, it is imperative to always use other, reliable sources to double-check the accuracy of any information generated by AI. This practice helps prevent the inclusion of 'AI hallucinations' in your academic work and ensures the integrity of your research and writing.